Date of Award

Spring 4-24-2020

Semester of Degree


Document Type

Open Access Dissertation

Degree Name

Ph.D. in Environmental Science


Forest and Natural Resources Management

Major Professor

Charles Kroll

Steering Committee Member

David Nowak

Steering Committee Member

Colin Beier

Steering Committee Member

Stephen Stehman

Steering Committee Member

John Drake

Steering Committee Member

John Wagner


Urban forest models can quantify forest structure and benefits, and are frequently employed in decision-making. This dissertation first reviewed case studies of urban forest modeling practices over the past two-decades, compared the similarities and differences among different models, and summarized the current trends and gaps in the field of urban forest modeling. One gap is the lack of uncertainty assessments for model output. To address this gap, this dissertation performed sensitivity and uncertainty analyses for a popular urban forest model, i-Tree Eco. Based on a case study in New York City, the sensitivity analyses found that the most important input variables are genus for isoprene and monoterpene emissions, DBH for carbon estimators, and leaf area index, temperature, and photosynthetically active radiation for dry deposition estimators. The uncertainty analyses addressed uncertainties associated with the entire i-Tree Eco modeling process, from input data collection, to the characterization of urban tree structure, to the subsequent estimators of the ecosystem services of urban trees. Uncertainty magnitudes were quantified by employing bootstrap and Monte Carlo simulations, and the three sources of uncertainty, input, model, and sampling, were aggregated to derive an estimator of total uncertainty. Through case studies in 16 cities across the United States, the average magnitude of total uncertainty across the 16 cities was 12.4% for leaf area, 12.4% for leaf biomass, 13.5% for carbon storage, 11.1% for carbon sequestration, 40.7% for isoprene emissions, and 25.0% for monoterpene emissions. For leaf and carbon estimators, the total uncertainty is primarily driven by sampling uncertainty, while the magnitudes of sampling, input and model uncertainty are similar across the 16 study cities. In contrast, input, sampling, and model uncertainties all contribute similarly to the total uncertainty for isoprene and monoterpene emission estimators, and there are larger variations in these three sources of uncertainty across the 16 study cities. To reduce overall uncertainty, future studies should develop more accurate urban-, local-, and species-specific allometric relationships, improve the spatial representation of meteorological variables, develop more extensive and accurate local-scale measurements to calibrate and verify models, and improve sampling strategies.